Automation is Not Always the Answer, in Retail or Healthcare
Lately, there have been many stories in the news about supermarkets ripping out self-service checkout scanner, stories like this: “Supermarkets start bagging self-serve checkouts.” 10 years ago, the supermarkets saw this technology as a clear cost savings – reducing labor costs. But, some costs increased, including “intentional and accidental theft, including misidentifying produce and baked goods as less-expensive varieties,” not to mention customers being slowed down or confused. Robots (like these scanners) can't be a smiling face… but I don't believe, for a minute, that the supermarkets are ripping the scanners out because they want to provide better customer service. I would guess that the cost savings didn't really materialize.
So what are the lessons and parallels for hospitals?
There are many cases where hospitals are seduced by technology, including:
- Robots that deliver meals or supplies to inpatient units
- Pharmacy robots that grab medications off a shelf to be delivered to patients
- “Track systems” that automate the movement of blood tubes in a laboratory
- Inventory cabinets that help track inventory and trigger restocking
- Self-service emergency department check in kiosks (a practice I was skeptical about 4 years ago)
What are my experiences with these technologies? Well, I've been hit by a delivery robot in a hallway after being promised “it had sensors that tell it to stop.”
I worked with a hospital pharmacy that realized the robot didn't really generate any labor savings because they had to pay somebody to run a machine that put pills into single-dose packaging that the robot could handle. When this pharmacy (at a children's hospital) built new space, they didn't buy a new robot.
I know of cases where hospital laboratories tore out their track systems (including this PDF case study from my former employer, ValuMetrix Services). Track systems are often called “total lab automation” and they are, from my experience, very slow conveyor belts that save labor but slow down turnaround times. These labs often automated a bad layout. When the high volume testing areas were really far away from the location where specimens arrived, the real Lean solution was changing the physical layout, not automating the waste of transportation. Washington Medical Center reduced FTEs when they put in the track system. When they took the track system out, under the guidance of a Lean consultant, they reduced turnaround times dramatically while adding back in just 1.5 FTEs. Sometimes, it's better to have people transport specimens instead of a robot.
Some hospitals have been ripping out automated inventory cabinets. I know of one that removed 93 out of 100 of these cabinets, replacing the cabinets and their expensive maintenance contracts with simple wire racks and two-bin kanban systems. When they looked deeply into their process, the hospital learned that a vast majority of “chargeable” supplies were not actually charged to specific patients. The billing functions that were tying usage into patient charts just weren't being used. The automated cabinets often created a bottleneck, where nurses had to stand in line to get supplies. With the open shelving, multiple nurses could grab what they needed at the same time, reducing delays in patient care. Automated cabinets were still retained when items were expensive and/or charged to patients and/or security threats.
I don't know if Parkland Hospital took out their E.D kiosk. If they follow the lead of supermarkets, they just might (or they already have).
I'm not anti-technology. I just believe strongly in the “Toyota Way” principle that states:
“Use Only Reliable, Thoroughly Tested Technology That Serves Your People and Processes”
In many of the cases above, technology was adding cost and delaying care… not really adding any value to the process.